To compare grey-matter (GM) and white-matter (WM) tau histopathology between PSP and Lewy body disorders (LBD: PDD and DLB) using digital histology.

Tau is the primary pathology in PSP and >50% of LBD have moderate/severe AD tau pathology at autopsy (LBD+tau).  In both disorders, tau pathology influences clinical features but few studies compare GM and WM pathologic distribution.

15 autopsy-confirmed PSP cases with parkinsonian presentations (PSP-RS:13, PSP-P:1, PSP-CBS:1) and 20 LBD+tau cases were selected (DLB:12, PDD:8). Angular gyrus (ANG), superior temporal gyrus (STG), and mid-frontal gyrus (MFG) were stained for tau (AT8). QuPath (v0.2.0-m2) calculated percentage area occupied (%AO) of tau inclusions in GM and WM. T-tests compared regional GM and WM tau%AO between and within LBD+tau and PSP. Linear regression assessed relationship between GM and WM tau%AO.

LBD+tau and PSP had equivalent GM tau%AO in all regions with trend-level elevations for LBD+tau in the STG (p=0.07).  PSP had higher WM tau%AO in all regions (p=0.005-0.03) except STG (p=0.37). PSP had higher WM/GM tau%AO ratios in all regions (p<0.001 for each).  Within LBD+tau, STG had the highest GM and WM tau%AO (GM: p<0.01 and WM: p=0.05  vs MFC and ANG).  Within PSP, cortical regions had equivalent GM and WM tau%AO (p>0.05).  GM tau%AO was significantly associated with WM tau%AO in both diseases (β=0.47-0.57, p<0.001); however, for each unit of GM tau%AO increase there was a significantly greater increase WM tau%AO in PSP than LBD+tau (p<0.001).

Using digital histology, we find novel postmortem evidence for divergent patterns of tau pathology in clinically similar Parkinsonian syndromes, with greater WM tau in PSP and equivalent GM cortical pathology compared with LBD with the possible exception of the temporal lobe where the greatest GM tau pathology was seen in LBD. These data suggest that tau-directed biomarker strategies with regional specificity may improve antemortem diagnosis.